A new hybrid learning-based algorithm for data clustering

摘要

In this paper a new hybrid algorithm based on particle swarm optimization (PSO), k-means and learning automata (KPSOLA) is proposed for data clustering. In the proposed algorithm, learning automata acts as the thinking brain of the particles in PSO. In each of iterations of the proposed algorithm execution, corresponding learning automata of each particle decides whether next move of that particle to be with respect to PSO algorithm or with respect to k-means algorithm. The proposed algorithm and also 4 other clustering algorithms have been used for clustering 6 standard datasets and their efficiencies are compared with each other. Experimental results show that the proposed algorithm has an acceptable efficiency and robustness.